We already have hydrogen bomb, why controllable nuclear fusion so difficult? And why controllable necessary? If we can detonate a tiny hydrogen bomb， we can collect the energy, like laser nuclear fusion do? Why didn't see any research of collider of coil gun? Beside LNF and Tokamak, using collide, we can let reaction happen in a space far from fragile device.
Why didn't see any research of collider of coil gun?
This was actually tried, although in a slightly different form. During WWII, while working at the Manhattan Project, James Tuck and Stanislaw Ulam used shaped charges to fire deuterium-infused metal foils at each other. No signs of fusion were seen.
The rough cross-section for the fusion reaction would have been known at that time, but not the scattering cross-section for the material itself. This turns out to be millions of times too high. In other words, it is extremely unlikely the fuel will ever get close enough to fuse.
In retrospect, this technique is good for creating reactions in the lab (ultimately it's similar to Rutherford's original experiment), but is many orders of magnitude away from the cross-section needed to produce positive energy output.
If we can detonate a tiny hydrogen bomb， we can collect the energy
We can. However, the minimum size of a fusion bomb is based on the minimum critical mass of the fission bomb used to ignite it. This means it is difficult to make a bomb smaller than about 100 T, and anything below about 10 kT tends to be extremely inefficient. So if you want to get a good energy bang for your plutonium buck, you have to use larger bombs, or else you may as well just burn the plutonium in a conventional fission reactor.
Nevertheless, some of these concepts were explored as part of PACER in the 1960s. However, after running for over a decade and resulting in a string of embarrassing failures, a 3rd party study considered the economics of the concept and rather conclusively showed that there was no way such a system could ever possibly be competitive with conventional fission, let alone other sources of energy. Funding was dropped in the 70s.
like laser nuclear fusion do
Ultimately these "ICF" devices get around the PACER problems by replacing the fission bomb with a powerful laser. Calculations showed there was no obvious lower-limit in that case, so you could make microbombs.
When Nuckolls was first studying them in the 1960s it seemed the laser would be less than 1 MJ for ignition, and one would get some reactions with drivers of a few kJ. In the 1970s we started building these kJ-sized machines, like Shiva, and as they ran it became clear the reaction rates were far below what they calculated.
A series of tests using x-rays released by a distant fission bomb were used to calibrate the calculations, and these suggested the amount of laser energy needed was about 100 MJ. We have no idea how to build such a device. LLNL argued that there were a number of reasons this was inaccurate and a "driver" around 2 MJ should do the job. So they built NIF, which delivers 4 MJ, to be on the safe side.
Well, it turns out experiments trump simulations once again, and it's pretty clear the original 100 MJ number is closer to the truth. Even with twice the calculated required energy, it appears highly unlikely NIF will reach "ignition". Even if it does, the expense of the laser, and it's horrible energy inefficiency, mean it could never be used as a power generator. If one could build a 100 MJ laser, the cost would be orders of magnitude higher than the economic value of the electricity it would produce.
You can read the whole story here.
reaction happen in a space
Fusion is already too expensive to contemplate, putting it in space won't help on that front!
That said, one of the most (oversold IMHO) advantages of fusion is its safety, so there's really no need to do this in the first place.